| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 400, | |
| "global_step": 468, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010683760683760684, | |
| "grad_norm": 14.031576787271042, | |
| "learning_rate": 5.106382978723404e-08, | |
| "logits/chosen": -1.478906273841858, | |
| "logits/rejected": -1.4328124523162842, | |
| "logps/chosen": -0.817187488079071, | |
| "logps/rejected": -0.850390613079071, | |
| "loss": 1.4433, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.4703125059604645, | |
| "rewards/chosen": -1.634374976158142, | |
| "rewards/margins": 0.06545410305261612, | |
| "rewards/rejected": -1.700781226158142, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021367521367521368, | |
| "grad_norm": 23.294649241820373, | |
| "learning_rate": 1.1489361702127659e-07, | |
| "logits/chosen": -1.4249999523162842, | |
| "logits/rejected": -1.388281226158142, | |
| "logps/chosen": -0.7289062738418579, | |
| "logps/rejected": -0.725781261920929, | |
| "loss": 1.4406, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.4859375059604645, | |
| "rewards/chosen": -1.4578125476837158, | |
| "rewards/margins": -0.00665283203125, | |
| "rewards/rejected": -1.451562523841858, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03205128205128205, | |
| "grad_norm": 10.86773840315628, | |
| "learning_rate": 1.7872340425531914e-07, | |
| "logits/chosen": -1.4929687976837158, | |
| "logits/rejected": -1.466406226158142, | |
| "logps/chosen": -0.7378906011581421, | |
| "logps/rejected": -0.7523437738418579, | |
| "loss": 1.3816, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -1.4757812023162842, | |
| "rewards/margins": 0.02743835374712944, | |
| "rewards/rejected": -1.5046875476837158, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042735042735042736, | |
| "grad_norm": 6.43024052602064, | |
| "learning_rate": 2.425531914893617e-07, | |
| "logits/chosen": -1.411718726158142, | |
| "logits/rejected": -1.3984375, | |
| "logps/chosen": -0.750781238079071, | |
| "logps/rejected": -0.848437488079071, | |
| "loss": 1.3784, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -1.501562476158142, | |
| "rewards/margins": 0.19492188096046448, | |
| "rewards/rejected": -1.696874976158142, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053418803418803416, | |
| "grad_norm": 11.19925501261024, | |
| "learning_rate": 3.063829787234042e-07, | |
| "logits/chosen": -1.4601562023162842, | |
| "logits/rejected": -1.4375, | |
| "logps/chosen": -0.7308593988418579, | |
| "logps/rejected": -0.778124988079071, | |
| "loss": 1.3567, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.4617187976837158, | |
| "rewards/margins": 0.09446106106042862, | |
| "rewards/rejected": -1.556249976158142, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0641025641025641, | |
| "grad_norm": 10.155982563226914, | |
| "learning_rate": 3.702127659574468e-07, | |
| "logits/chosen": -1.4421875476837158, | |
| "logits/rejected": -1.4148437976837158, | |
| "logps/chosen": -0.6324218511581421, | |
| "logps/rejected": -0.66796875, | |
| "loss": 1.3238, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.4906249940395355, | |
| "rewards/chosen": -1.2648437023162842, | |
| "rewards/margins": 0.07158203423023224, | |
| "rewards/rejected": -1.3359375, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07478632478632478, | |
| "grad_norm": 12.236975012364756, | |
| "learning_rate": 4.3404255319148936e-07, | |
| "logits/chosen": -1.4679687023162842, | |
| "logits/rejected": -1.4406249523162842, | |
| "logps/chosen": -0.583203136920929, | |
| "logps/rejected": -0.599609375, | |
| "loss": 1.3457, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -1.166406273841858, | |
| "rewards/margins": 0.03232421725988388, | |
| "rewards/rejected": -1.19921875, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08547008547008547, | |
| "grad_norm": 15.99115276248814, | |
| "learning_rate": 4.978723404255318e-07, | |
| "logits/chosen": -1.497656226158142, | |
| "logits/rejected": -1.4500000476837158, | |
| "logps/chosen": -0.5269531011581421, | |
| "logps/rejected": -0.552734375, | |
| "loss": 1.3196, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.4984374940395355, | |
| "rewards/chosen": -1.0539062023162842, | |
| "rewards/margins": 0.05270995944738388, | |
| "rewards/rejected": -1.10546875, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09615384615384616, | |
| "grad_norm": 6.342347202429818, | |
| "learning_rate": 5.617021276595744e-07, | |
| "logits/chosen": -1.540624976158142, | |
| "logits/rejected": -1.510156273841858, | |
| "logps/chosen": -0.4320312440395355, | |
| "logps/rejected": -0.46113282442092896, | |
| "loss": 1.2917, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5171874761581421, | |
| "rewards/chosen": -0.864062488079071, | |
| "rewards/margins": 0.05752868577837944, | |
| "rewards/rejected": -0.9222656488418579, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10683760683760683, | |
| "grad_norm": 6.093826949661445, | |
| "learning_rate": 5.9996658982558e-07, | |
| "logits/chosen": -1.5, | |
| "logits/rejected": -1.471093773841858, | |
| "logps/chosen": -0.3994140625, | |
| "logps/rejected": -0.43671876192092896, | |
| "loss": 1.2759, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -0.798828125, | |
| "rewards/margins": 0.07478027045726776, | |
| "rewards/rejected": -0.8734375238418579, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11752136752136752, | |
| "grad_norm": 32.342746515822284, | |
| "learning_rate": 5.995908108202275e-07, | |
| "logits/chosen": -1.4992187023162842, | |
| "logits/rejected": -1.470312476158142, | |
| "logps/chosen": -0.3929687440395355, | |
| "logps/rejected": -0.42460936307907104, | |
| "loss": 1.2828, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -0.785937488079071, | |
| "rewards/margins": 0.06329345703125, | |
| "rewards/rejected": -0.8492187261581421, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282051282051282, | |
| "grad_norm": 7.8916210409345195, | |
| "learning_rate": 5.98798014903091e-07, | |
| "logits/chosen": -1.5226562023162842, | |
| "logits/rejected": -1.4874999523162842, | |
| "logps/chosen": -0.40058594942092896, | |
| "logps/rejected": -0.4244140684604645, | |
| "loss": 1.2934, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.526562511920929, | |
| "rewards/chosen": -0.8011718988418579, | |
| "rewards/margins": 0.04749755933880806, | |
| "rewards/rejected": -0.848828136920929, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 10.591326680588619, | |
| "learning_rate": 5.97589305610371e-07, | |
| "logits/chosen": -1.5109374523162842, | |
| "logits/rejected": -1.467187523841858, | |
| "logps/chosen": -0.3916015625, | |
| "logps/rejected": -0.4154296815395355, | |
| "loss": 1.2916, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.783203125, | |
| "rewards/margins": 0.04823913425207138, | |
| "rewards/rejected": -0.830859363079071, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14957264957264957, | |
| "grad_norm": 14.122292943634292, | |
| "learning_rate": 5.959663654109491e-07, | |
| "logits/chosen": -1.517187476158142, | |
| "logits/rejected": -1.478906273841858, | |
| "logps/chosen": -0.40234375, | |
| "logps/rejected": -0.4400390684604645, | |
| "loss": 1.2902, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.8046875, | |
| "rewards/margins": 0.07520751655101776, | |
| "rewards/rejected": -0.880078136920929, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16025641025641027, | |
| "grad_norm": 11.624864350308169, | |
| "learning_rate": 5.93931453364468e-07, | |
| "logits/chosen": -1.5617187023162842, | |
| "logits/rejected": -1.54296875, | |
| "logps/chosen": -0.40234375, | |
| "logps/rejected": -0.45429688692092896, | |
| "loss": 1.2626, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5718749761581421, | |
| "rewards/chosen": -0.8046875, | |
| "rewards/margins": 0.10439453274011612, | |
| "rewards/rejected": -0.9085937738418579, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 8.890913992364057, | |
| "learning_rate": 5.914874019768214e-07, | |
| "logits/chosen": -1.575781226158142, | |
| "logits/rejected": -1.5593750476837158, | |
| "logps/chosen": -0.4203124940395355, | |
| "logps/rejected": -0.4765625, | |
| "loss": 1.265, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.840624988079071, | |
| "rewards/margins": 0.11260070651769638, | |
| "rewards/rejected": -0.953125, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18162393162393162, | |
| "grad_norm": 9.053037451143402, | |
| "learning_rate": 5.886376132574339e-07, | |
| "logits/chosen": -1.517968773841858, | |
| "logits/rejected": -1.489843726158142, | |
| "logps/chosen": -0.44257813692092896, | |
| "logps/rejected": -0.5015624761581421, | |
| "loss": 1.2794, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5171874761581421, | |
| "rewards/chosen": -0.8851562738418579, | |
| "rewards/margins": 0.11624755710363388, | |
| "rewards/rejected": -1.0031249523162842, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19230769230769232, | |
| "grad_norm": 7.52173535900888, | |
| "learning_rate": 5.853860539838169e-07, | |
| "logits/chosen": -1.5085937976837158, | |
| "logits/rejected": -1.4734375476837158, | |
| "logps/chosen": -0.4273437559604645, | |
| "logps/rejected": -0.4761718809604645, | |
| "loss": 1.2881, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.528124988079071, | |
| "rewards/chosen": -0.854687511920929, | |
| "rewards/margins": 0.09850464016199112, | |
| "rewards/rejected": -0.952343761920929, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.202991452991453, | |
| "grad_norm": 7.370453197913243, | |
| "learning_rate": 5.817372501799931e-07, | |
| "logits/chosen": -1.55859375, | |
| "logits/rejected": -1.5499999523162842, | |
| "logps/chosen": -0.38457030057907104, | |
| "logps/rejected": -0.44902342557907104, | |
| "loss": 1.2621, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.7691406011581421, | |
| "rewards/margins": 0.12803955376148224, | |
| "rewards/rejected": -0.8980468511581421, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21367521367521367, | |
| "grad_norm": 13.697273867527963, | |
| "learning_rate": 5.776962808164751e-07, | |
| "logits/chosen": -1.560156226158142, | |
| "logits/rejected": -1.513281226158142, | |
| "logps/chosen": -0.39531248807907104, | |
| "logps/rejected": -0.4423828125, | |
| "loss": 1.2896, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -0.7906249761581421, | |
| "rewards/margins": 0.09405517578125, | |
| "rewards/rejected": -0.884765625, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22435897435897437, | |
| "grad_norm": 14.90977337445018, | |
| "learning_rate": 5.73268770740567e-07, | |
| "logits/chosen": -1.521875023841858, | |
| "logits/rejected": -1.4874999523162842, | |
| "logps/chosen": -0.38300782442092896, | |
| "logps/rejected": -0.43476563692092896, | |
| "loss": 1.275, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.526562511920929, | |
| "rewards/chosen": -0.7660156488418579, | |
| "rewards/margins": 0.10239257663488388, | |
| "rewards/rejected": -0.8695312738418579, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23504273504273504, | |
| "grad_norm": 12.70136475909842, | |
| "learning_rate": 5.684608828468305e-07, | |
| "logits/chosen": -1.544531226158142, | |
| "logits/rejected": -1.501562476158142, | |
| "logps/chosen": -0.3892578184604645, | |
| "logps/rejected": -0.42597657442092896, | |
| "loss": 1.2849, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.778515636920929, | |
| "rewards/margins": 0.07318115234375, | |
| "rewards/rejected": -0.8519531488418579, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24572649572649571, | |
| "grad_norm": 10.809074159755266, | |
| "learning_rate": 5.632793094986134e-07, | |
| "logits/chosen": -1.5304687023162842, | |
| "logits/rejected": -1.5109374523162842, | |
| "logps/chosen": -0.39667969942092896, | |
| "logps/rejected": -0.4609375, | |
| "loss": 1.2658, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5484374761581421, | |
| "rewards/chosen": -0.7933593988418579, | |
| "rewards/margins": 0.12885741889476776, | |
| "rewards/rejected": -0.921875, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 7.310341755161344, | |
| "learning_rate": 5.577312632125811e-07, | |
| "logits/chosen": -1.5437500476837158, | |
| "logits/rejected": -1.532812476158142, | |
| "logps/chosen": -0.40058594942092896, | |
| "logps/rejected": -0.4585937559604645, | |
| "loss": 1.261, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5640624761581421, | |
| "rewards/chosen": -0.8011718988418579, | |
| "rewards/margins": 0.11567382514476776, | |
| "rewards/rejected": -0.917187511920929, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2670940170940171, | |
| "grad_norm": 8.26904210261963, | |
| "learning_rate": 5.518244666192185e-07, | |
| "logits/chosen": -1.552343726158142, | |
| "logits/rejected": -1.5281250476837158, | |
| "logps/chosen": -0.45878905057907104, | |
| "logps/rejected": -0.5210937261581421, | |
| "loss": 1.2867, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.9175781011581421, | |
| "rewards/margins": 0.12409820407629013, | |
| "rewards/rejected": -1.0421874523162842, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 10.250708094357067, | |
| "learning_rate": 5.455671417132762e-07, | |
| "logits/chosen": -1.5671875476837158, | |
| "logits/rejected": -1.5164062976837158, | |
| "logps/chosen": -0.40839844942092896, | |
| "logps/rejected": -0.49785155057907104, | |
| "loss": 1.2559, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.526562511920929, | |
| "rewards/chosen": -0.8167968988418579, | |
| "rewards/margins": 0.18012695014476776, | |
| "rewards/rejected": -0.9957031011581421, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28846153846153844, | |
| "grad_norm": 9.573630775951374, | |
| "learning_rate": 5.389679984091244e-07, | |
| "logits/chosen": -1.5554687976837158, | |
| "logits/rejected": -1.521093726158142, | |
| "logps/chosen": -0.40156251192092896, | |
| "logps/rejected": -0.4853515625, | |
| "loss": 1.2483, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5484374761581421, | |
| "rewards/chosen": -0.8031250238418579, | |
| "rewards/margins": 0.16811522841453552, | |
| "rewards/rejected": -0.970703125, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29914529914529914, | |
| "grad_norm": 15.240586201261362, | |
| "learning_rate": 5.320362224169441e-07, | |
| "logits/chosen": -1.5382812023162842, | |
| "logits/rejected": -1.5125000476837158, | |
| "logps/chosen": -0.4029296934604645, | |
| "logps/rejected": -0.4652343690395355, | |
| "loss": 1.2741, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5640624761581421, | |
| "rewards/chosen": -0.805859386920929, | |
| "rewards/margins": 0.12409667670726776, | |
| "rewards/rejected": -0.930468738079071, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30982905982905984, | |
| "grad_norm": 11.28111450928389, | |
| "learning_rate": 5.247814624566326e-07, | |
| "logits/chosen": -1.525781273841858, | |
| "logits/rejected": -1.49609375, | |
| "logps/chosen": -0.3916015625, | |
| "logps/rejected": -0.46757811307907104, | |
| "loss": 1.2728, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.783203125, | |
| "rewards/margins": 0.15163573622703552, | |
| "rewards/rejected": -0.9351562261581421, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32051282051282054, | |
| "grad_norm": 13.169292810618147, | |
| "learning_rate": 5.172138168272204e-07, | |
| "logits/chosen": -1.5265624523162842, | |
| "logits/rejected": -1.497656226158142, | |
| "logps/chosen": -0.40546876192092896, | |
| "logps/rejected": -0.4759765565395355, | |
| "loss": 1.2561, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.559374988079071, | |
| "rewards/chosen": -0.8109375238418579, | |
| "rewards/margins": 0.13974609971046448, | |
| "rewards/rejected": -0.951953113079071, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3311965811965812, | |
| "grad_norm": 10.242809753624087, | |
| "learning_rate": 5.093438193504931e-07, | |
| "logits/chosen": -1.53125, | |
| "logits/rejected": -1.4914062023162842, | |
| "logps/chosen": -0.38749998807907104, | |
| "logps/rejected": -0.4957031309604645, | |
| "loss": 1.234, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5406249761581421, | |
| "rewards/chosen": -0.7749999761581421, | |
| "rewards/margins": 0.21766357123851776, | |
| "rewards/rejected": -0.991406261920929, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 10.439751676379643, | |
| "learning_rate": 5.011824247083868e-07, | |
| "logits/chosen": -1.52734375, | |
| "logits/rejected": -1.4968750476837158, | |
| "logps/chosen": -0.4423828125, | |
| "logps/rejected": -0.5267578363418579, | |
| "loss": 1.2656, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.884765625, | |
| "rewards/margins": 0.16884765028953552, | |
| "rewards/rejected": -1.0535156726837158, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3525641025641026, | |
| "grad_norm": 14.3284012618235, | |
| "learning_rate": 4.927409931945641e-07, | |
| "logits/chosen": -1.5789062976837158, | |
| "logits/rejected": -1.529687523841858, | |
| "logps/chosen": -0.4091796875, | |
| "logps/rejected": -0.5054687261581421, | |
| "loss": 1.2416, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.818359375, | |
| "rewards/margins": 0.19272461533546448, | |
| "rewards/rejected": -1.0109374523162842, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.36324786324786323, | |
| "grad_norm": 14.80755642701636, | |
| "learning_rate": 4.840312749013969e-07, | |
| "logits/chosen": -1.5320312976837158, | |
| "logits/rejected": -1.502343773841858, | |
| "logps/chosen": -0.4046874940395355, | |
| "logps/rejected": -0.508984386920929, | |
| "loss": 1.2361, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.809374988079071, | |
| "rewards/margins": 0.20823974907398224, | |
| "rewards/rejected": -1.017968773841858, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37393162393162394, | |
| "grad_norm": 14.25663733412322, | |
| "learning_rate": 4.750653933643674e-07, | |
| "logits/chosen": -1.529687523841858, | |
| "logits/rejected": -1.5046875476837158, | |
| "logps/chosen": -0.4306640625, | |
| "logps/rejected": -0.591015636920929, | |
| "loss": 1.2129, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.559374988079071, | |
| "rewards/chosen": -0.861328125, | |
| "rewards/margins": 0.3213867247104645, | |
| "rewards/rejected": -1.182031273841858, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 15.795370678351055, | |
| "learning_rate": 4.6585582868665284e-07, | |
| "logits/chosen": -1.564062476158142, | |
| "logits/rejected": -1.5281250476837158, | |
| "logps/chosen": -0.42753905057907104, | |
| "logps/rejected": -0.5181640386581421, | |
| "loss": 1.252, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5640624761581421, | |
| "rewards/chosen": -0.8550781011581421, | |
| "rewards/margins": 0.18134765326976776, | |
| "rewards/rejected": -1.0363280773162842, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3952991452991453, | |
| "grad_norm": 14.851431627716563, | |
| "learning_rate": 4.5641540016738414e-07, | |
| "logits/chosen": -1.5078125, | |
| "logits/rejected": -1.486718773841858, | |
| "logps/chosen": -0.4833984375, | |
| "logps/rejected": -0.573437511920929, | |
| "loss": 1.2683, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5296875238418579, | |
| "rewards/chosen": -0.966796875, | |
| "rewards/margins": 0.18000487983226776, | |
| "rewards/rejected": -1.146875023841858, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.405982905982906, | |
| "grad_norm": 16.66852729151085, | |
| "learning_rate": 4.467572484577589e-07, | |
| "logits/chosen": -1.5695312023162842, | |
| "logits/rejected": -1.5398437976837158, | |
| "logps/chosen": -0.4722656309604645, | |
| "logps/rejected": -0.636523425579071, | |
| "loss": 1.2018, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5843750238418579, | |
| "rewards/chosen": -0.944531261920929, | |
| "rewards/margins": 0.327392578125, | |
| "rewards/rejected": -1.273046851158142, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 15.67537550174711, | |
| "learning_rate": 4.368948172698474e-07, | |
| "logits/chosen": -1.5234375, | |
| "logits/rejected": -1.4929687976837158, | |
| "logps/chosen": -0.47441405057907104, | |
| "logps/rejected": -0.5902343988418579, | |
| "loss": 1.2406, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5171874761581421, | |
| "rewards/chosen": -0.9488281011581421, | |
| "rewards/margins": 0.23137207329273224, | |
| "rewards/rejected": -1.1804687976837158, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "grad_norm": 15.569695653680345, | |
| "learning_rate": 4.2684183466355006e-07, | |
| "logits/chosen": -1.524999976158142, | |
| "logits/rejected": -1.486718773841858, | |
| "logps/chosen": -0.4554687440395355, | |
| "logps/rejected": -0.570117175579071, | |
| "loss": 1.2451, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5484374761581421, | |
| "rewards/chosen": -0.910937488079071, | |
| "rewards/margins": 0.22930297255516052, | |
| "rewards/rejected": -1.140234351158142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43803418803418803, | |
| "grad_norm": 15.479616500843248, | |
| "learning_rate": 4.1661229393775783e-07, | |
| "logits/chosen": -1.5539062023162842, | |
| "logits/rejected": -1.528906226158142, | |
| "logps/chosen": -0.43828123807907104, | |
| "logps/rejected": -0.5511718988418579, | |
| "loss": 1.2274, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.8765624761581421, | |
| "rewards/margins": 0.22568359971046448, | |
| "rewards/rejected": -1.1023437976837158, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44871794871794873, | |
| "grad_norm": 14.30449666703877, | |
| "learning_rate": 4.062204341523094e-07, | |
| "logits/chosen": -1.5859375, | |
| "logits/rejected": -1.5390625, | |
| "logps/chosen": -0.530468761920929, | |
| "logps/rejected": -0.673828125, | |
| "loss": 1.25, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5609375238418579, | |
| "rewards/chosen": -1.060937523841858, | |
| "rewards/margins": 0.2860351502895355, | |
| "rewards/rejected": -1.34765625, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4594017094017094, | |
| "grad_norm": 12.308222317827216, | |
| "learning_rate": 3.9568072030786214e-07, | |
| "logits/chosen": -1.548437476158142, | |
| "logits/rejected": -1.5148437023162842, | |
| "logps/chosen": -0.5416015386581421, | |
| "logps/rejected": -0.7269531488418579, | |
| "loss": 1.2141, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5796874761581421, | |
| "rewards/chosen": -1.0832030773162842, | |
| "rewards/margins": 0.3740234375, | |
| "rewards/rejected": -1.4539062976837158, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4700854700854701, | |
| "grad_norm": 17.12922996772855, | |
| "learning_rate": 3.8500782321126227e-07, | |
| "logits/chosen": -1.5398437976837158, | |
| "logits/rejected": -1.501562476158142, | |
| "logps/chosen": -0.559374988079071, | |
| "logps/rejected": -0.66015625, | |
| "loss": 1.2812, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5640624761581421, | |
| "rewards/chosen": -1.118749976158142, | |
| "rewards/margins": 0.20118407905101776, | |
| "rewards/rejected": -1.3203125, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4807692307692308, | |
| "grad_norm": 11.416683434770386, | |
| "learning_rate": 3.742165990544432e-07, | |
| "logits/chosen": -1.49609375, | |
| "logits/rejected": -1.46484375, | |
| "logps/chosen": -0.4810546934604645, | |
| "logps/rejected": -0.5923827886581421, | |
| "loss": 1.2637, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5718749761581421, | |
| "rewards/chosen": -0.962109386920929, | |
| "rewards/margins": 0.22169189155101776, | |
| "rewards/rejected": -1.1847655773162842, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49145299145299143, | |
| "grad_norm": 11.009026957181359, | |
| "learning_rate": 3.633220687352755e-07, | |
| "logits/chosen": -1.528906226158142, | |
| "logits/rejected": -1.506250023841858, | |
| "logps/chosen": -0.451171875, | |
| "logps/rejected": -0.558789074420929, | |
| "loss": 1.2326, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.90234375, | |
| "rewards/margins": 0.21640625596046448, | |
| "rewards/rejected": -1.117578148841858, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5021367521367521, | |
| "grad_norm": 9.790833316230884, | |
| "learning_rate": 3.523393969491533e-07, | |
| "logits/chosen": -1.537500023841858, | |
| "logits/rejected": -1.49609375, | |
| "logps/chosen": -0.44843751192092896, | |
| "logps/rejected": -0.553906261920929, | |
| "loss": 1.2396, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.8968750238418579, | |
| "rewards/margins": 0.21157225966453552, | |
| "rewards/rejected": -1.107812523841858, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 11.337317668384426, | |
| "learning_rate": 3.4128387108042213e-07, | |
| "logits/chosen": -1.521093726158142, | |
| "logits/rejected": -1.5046875476837158, | |
| "logps/chosen": -0.47050780057907104, | |
| "logps/rejected": -0.603320300579071, | |
| "loss": 1.2046, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9410156011581421, | |
| "rewards/margins": 0.2652343809604645, | |
| "rewards/rejected": -1.206640601158142, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5235042735042735, | |
| "grad_norm": 7.811613503865328, | |
| "learning_rate": 3.301708799230284e-07, | |
| "logits/chosen": -1.548437476158142, | |
| "logits/rejected": -1.5031249523162842, | |
| "logps/chosen": -0.484375, | |
| "logps/rejected": -0.628125011920929, | |
| "loss": 1.2204, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5765625238418579, | |
| "rewards/chosen": -0.96875, | |
| "rewards/margins": 0.2870117127895355, | |
| "rewards/rejected": -1.256250023841858, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5341880341880342, | |
| "grad_norm": 12.75427190985336, | |
| "learning_rate": 3.1901589226001136e-07, | |
| "logits/chosen": -1.560156226158142, | |
| "logits/rejected": -1.5242187976837158, | |
| "logps/chosen": -0.509570300579071, | |
| "logps/rejected": -0.595703125, | |
| "loss": 1.2453, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -1.019140601158142, | |
| "rewards/margins": 0.17375488579273224, | |
| "rewards/rejected": -1.19140625, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5448717948717948, | |
| "grad_norm": 14.20386765941155, | |
| "learning_rate": 3.0783443533165413e-07, | |
| "logits/chosen": -1.5554687976837158, | |
| "logits/rejected": -1.5226562023162842, | |
| "logps/chosen": -0.501953125, | |
| "logps/rejected": -0.643359363079071, | |
| "loss": 1.2107, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5718749761581421, | |
| "rewards/chosen": -1.00390625, | |
| "rewards/margins": 0.283447265625, | |
| "rewards/rejected": -1.286718726158142, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 12.99228285452067, | |
| "learning_rate": 2.9664207322226467e-07, | |
| "logits/chosen": -1.55078125, | |
| "logits/rejected": -1.533593773841858, | |
| "logps/chosen": -0.5103515386581421, | |
| "logps/rejected": -0.6527343988418579, | |
| "loss": 1.2294, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.598437488079071, | |
| "rewards/chosen": -1.0207030773162842, | |
| "rewards/margins": 0.2850585877895355, | |
| "rewards/rejected": -1.3054687976837158, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5662393162393162, | |
| "grad_norm": 13.73274782795775, | |
| "learning_rate": 2.8545438519567173e-07, | |
| "logits/chosen": -1.521875023841858, | |
| "logits/rejected": -1.4968750476837158, | |
| "logps/chosen": -0.47832030057907104, | |
| "logps/rejected": -0.582226574420929, | |
| "loss": 1.2485, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.567187488079071, | |
| "rewards/chosen": -0.9566406011581421, | |
| "rewards/margins": 0.20791015028953552, | |
| "rewards/rejected": -1.164453148841858, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5769230769230769, | |
| "grad_norm": 12.89822692411219, | |
| "learning_rate": 2.742869440095914e-07, | |
| "logits/chosen": -1.53125, | |
| "logits/rejected": -1.505468726158142, | |
| "logps/chosen": -0.5, | |
| "logps/rejected": -0.6136718988418579, | |
| "loss": 1.2508, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5546875, | |
| "rewards/chosen": -1.0, | |
| "rewards/margins": 0.22727051377296448, | |
| "rewards/rejected": -1.2273437976837158, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5876068376068376, | |
| "grad_norm": 15.122528311539122, | |
| "learning_rate": 2.6315529423904975e-07, | |
| "logits/chosen": -1.556249976158142, | |
| "logits/rejected": -1.5242187976837158, | |
| "logps/chosen": -0.48945313692092896, | |
| "logps/rejected": -0.6246093511581421, | |
| "loss": 1.2162, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6031249761581421, | |
| "rewards/chosen": -0.9789062738418579, | |
| "rewards/margins": 0.2700439393520355, | |
| "rewards/rejected": -1.2492187023162842, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5982905982905983, | |
| "grad_norm": 10.081247455678685, | |
| "learning_rate": 2.5207493063903506e-07, | |
| "logits/chosen": -1.524999976158142, | |
| "logits/rejected": -1.4874999523162842, | |
| "logps/chosen": -0.48261719942092896, | |
| "logps/rejected": -0.626953125, | |
| "loss": 1.2118, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -0.9652343988418579, | |
| "rewards/margins": 0.28764647245407104, | |
| "rewards/rejected": -1.25390625, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6089743589743589, | |
| "grad_norm": 10.004059372308623, | |
| "learning_rate": 2.410612765764965e-07, | |
| "logits/chosen": -1.5476562976837158, | |
| "logits/rejected": -1.493749976158142, | |
| "logps/chosen": -0.508593738079071, | |
| "logps/rejected": -0.637499988079071, | |
| "loss": 1.2313, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -1.017187476158142, | |
| "rewards/margins": 0.2581420838832855, | |
| "rewards/rejected": -1.274999976158142, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6196581196581197, | |
| "grad_norm": 13.343918364038107, | |
| "learning_rate": 2.301296625617126e-07, | |
| "logits/chosen": -1.56640625, | |
| "logits/rejected": -1.5421874523162842, | |
| "logps/chosen": -0.5296875238418579, | |
| "logps/rejected": -0.6832031011581421, | |
| "loss": 1.2206, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5953124761581421, | |
| "rewards/chosen": -1.0593750476837158, | |
| "rewards/margins": 0.30791014432907104, | |
| "rewards/rejected": -1.3664062023162842, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6303418803418803, | |
| "grad_norm": 11.405690420163953, | |
| "learning_rate": 2.1929530490891143e-07, | |
| "logits/chosen": -1.54296875, | |
| "logits/rejected": -1.5203125476837158, | |
| "logps/chosen": -0.513867199420929, | |
| "logps/rejected": -0.6363281011581421, | |
| "loss": 1.24, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5796874761581421, | |
| "rewards/chosen": -1.027734398841858, | |
| "rewards/margins": 0.24458007514476776, | |
| "rewards/rejected": -1.2726562023162842, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6410256410256411, | |
| "grad_norm": 11.838110328694215, | |
| "learning_rate": 2.0857328455584657e-07, | |
| "logits/chosen": -1.5398437976837158, | |
| "logits/rejected": -1.514062523841858, | |
| "logps/chosen": -0.5025390386581421, | |
| "logps/rejected": -0.637499988079071, | |
| "loss": 1.2237, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5843750238418579, | |
| "rewards/chosen": -1.0050780773162842, | |
| "rewards/margins": 0.270263671875, | |
| "rewards/rejected": -1.274999976158142, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6517094017094017, | |
| "grad_norm": 9.470686552249557, | |
| "learning_rate": 1.9797852607181124e-07, | |
| "logits/chosen": -1.576562523841858, | |
| "logits/rejected": -1.545312523841858, | |
| "logps/chosen": -0.47382813692092896, | |
| "logps/rejected": -0.627734363079071, | |
| "loss": 1.2092, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5921875238418579, | |
| "rewards/chosen": -0.9476562738418579, | |
| "rewards/margins": 0.3075195252895355, | |
| "rewards/rejected": -1.255468726158142, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6623931623931624, | |
| "grad_norm": 28.661639533450387, | |
| "learning_rate": 1.875257768833105e-07, | |
| "logits/chosen": -1.509374976158142, | |
| "logits/rejected": -1.493749976158142, | |
| "logps/chosen": -0.5083984136581421, | |
| "logps/rejected": -0.6343749761581421, | |
| "loss": 1.2552, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -1.0167968273162842, | |
| "rewards/margins": 0.25190430879592896, | |
| "rewards/rejected": -1.2687499523162842, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6730769230769231, | |
| "grad_norm": 12.529058191151396, | |
| "learning_rate": 1.7722958674630727e-07, | |
| "logits/chosen": -1.5234375, | |
| "logits/rejected": -1.5031249523162842, | |
| "logps/chosen": -0.521289050579071, | |
| "logps/rejected": -0.6499999761581421, | |
| "loss": 1.2271, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.042578101158142, | |
| "rewards/margins": 0.2585205137729645, | |
| "rewards/rejected": -1.2999999523162842, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 13.436410257147102, | |
| "learning_rate": 1.6710428749361832e-07, | |
| "logits/chosen": -1.525781273841858, | |
| "logits/rejected": -1.490625023841858, | |
| "logps/chosen": -0.5423828363418579, | |
| "logps/rejected": -0.701953113079071, | |
| "loss": 1.2103, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0847656726837158, | |
| "rewards/margins": 0.3174804747104645, | |
| "rewards/rejected": -1.403906226158142, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6944444444444444, | |
| "grad_norm": 15.554250562987784, | |
| "learning_rate": 1.571639730856482e-07, | |
| "logits/chosen": -1.5203125476837158, | |
| "logits/rejected": -1.4914062023162842, | |
| "logps/chosen": -0.5386718511581421, | |
| "logps/rejected": -0.6773437261581421, | |
| "loss": 1.2158, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.596875011920929, | |
| "rewards/chosen": -1.0773437023162842, | |
| "rewards/margins": 0.27617186307907104, | |
| "rewards/rejected": -1.3546874523162842, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7051282051282052, | |
| "grad_norm": 11.275430224004403, | |
| "learning_rate": 1.4742247999223112e-07, | |
| "logits/chosen": -1.5398437976837158, | |
| "logits/rejected": -1.51953125, | |
| "logps/chosen": -0.5498046875, | |
| "logps/rejected": -0.710156261920929, | |
| "loss": 1.2059, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5953124761581421, | |
| "rewards/chosen": -1.099609375, | |
| "rewards/margins": 0.3193359375, | |
| "rewards/rejected": -1.420312523841858, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7158119658119658, | |
| "grad_norm": 10.324498809414596, | |
| "learning_rate": 1.3789336793288877e-07, | |
| "logits/chosen": -1.489843726158142, | |
| "logits/rejected": -1.482031226158142, | |
| "logps/chosen": -0.538281261920929, | |
| "logps/rejected": -0.717578113079071, | |
| "loss": 1.187, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6078125238418579, | |
| "rewards/chosen": -1.076562523841858, | |
| "rewards/margins": 0.357666015625, | |
| "rewards/rejected": -1.435156226158142, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7264957264957265, | |
| "grad_norm": 14.78536991969488, | |
| "learning_rate": 1.285899010023115e-07, | |
| "logits/chosen": -1.5421874523162842, | |
| "logits/rejected": -1.517187476158142, | |
| "logps/chosen": -0.540820300579071, | |
| "logps/rejected": -0.678515613079071, | |
| "loss": 1.2288, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5921875238418579, | |
| "rewards/chosen": -1.081640601158142, | |
| "rewards/margins": 0.2734375, | |
| "rewards/rejected": -1.357031226158142, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7371794871794872, | |
| "grad_norm": 11.083012202046339, | |
| "learning_rate": 1.1952502920733554e-07, | |
| "logits/chosen": -1.509374976158142, | |
| "logits/rejected": -1.4968750476837158, | |
| "logps/chosen": -0.49921876192092896, | |
| "logps/rejected": -0.645703136920929, | |
| "loss": 1.2018, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.9984375238418579, | |
| "rewards/margins": 0.2920898497104645, | |
| "rewards/rejected": -1.291406273841858, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7478632478632479, | |
| "grad_norm": 13.711430399043142, | |
| "learning_rate": 1.1071137044111784e-07, | |
| "logits/chosen": -1.514062523841858, | |
| "logits/rejected": -1.4921875, | |
| "logps/chosen": -0.48417967557907104, | |
| "logps/rejected": -0.6484375, | |
| "loss": 1.2034, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -0.9683593511581421, | |
| "rewards/margins": 0.3277343809604645, | |
| "rewards/rejected": -1.296875, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7585470085470085, | |
| "grad_norm": 17.091136083884738, | |
| "learning_rate": 1.0216119291959624e-07, | |
| "logits/chosen": -1.5234375, | |
| "logits/rejected": -1.4874999523162842, | |
| "logps/chosen": -0.5474609136581421, | |
| "logps/rejected": -0.705859363079071, | |
| "loss": 1.2126, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.589062511920929, | |
| "rewards/chosen": -1.0949218273162842, | |
| "rewards/margins": 0.3169921934604645, | |
| "rewards/rejected": -1.411718726158142, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 11.69714980581959, | |
| "learning_rate": 9.388639810468644e-08, | |
| "logits/chosen": -1.528906226158142, | |
| "logits/rejected": -1.498437523841858, | |
| "logps/chosen": -0.516406238079071, | |
| "logps/rejected": -0.6910156011581421, | |
| "loss": 1.1951, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.573437511920929, | |
| "rewards/chosen": -1.032812476158142, | |
| "rewards/margins": 0.3477539122104645, | |
| "rewards/rejected": -1.3820312023162842, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7799145299145299, | |
| "grad_norm": 9.769420151980366, | |
| "learning_rate": 8.589850413798264e-08, | |
| "logits/chosen": -1.5578124523162842, | |
| "logits/rejected": -1.5226562023162842, | |
| "logps/chosen": -0.513476550579071, | |
| "logps/rejected": -0.6685546636581421, | |
| "loss": 1.1922, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.026953101158142, | |
| "rewards/margins": 0.3106933534145355, | |
| "rewards/rejected": -1.3371093273162842, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7905982905982906, | |
| "grad_norm": 16.620091820620416, | |
| "learning_rate": 7.82086298080236e-08, | |
| "logits/chosen": -1.514062523841858, | |
| "logits/rejected": -1.474218726158142, | |
| "logps/chosen": -0.5093749761581421, | |
| "logps/rejected": -0.6572265625, | |
| "loss": 1.2093, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0187499523162842, | |
| "rewards/margins": 0.29389649629592896, | |
| "rewards/rejected": -1.314453125, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8012820512820513, | |
| "grad_norm": 13.721108190371638, | |
| "learning_rate": 7.082747907343972e-08, | |
| "logits/chosen": -1.51171875, | |
| "logits/rejected": -1.484375, | |
| "logps/chosen": -0.528515636920929, | |
| "logps/rejected": -0.6703125238418579, | |
| "loss": 1.2263, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.057031273841858, | |
| "rewards/margins": 0.283203125, | |
| "rewards/rejected": -1.3406250476837158, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.811965811965812, | |
| "grad_norm": 10.832579603627439, | |
| "learning_rate": 6.37653261635252e-08, | |
| "logits/chosen": -1.560156226158142, | |
| "logits/rejected": -1.524999976158142, | |
| "logps/chosen": -0.5396484136581421, | |
| "logps/rejected": -0.745312511920929, | |
| "loss": 1.1876, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5921875238418579, | |
| "rewards/chosen": -1.0792968273162842, | |
| "rewards/margins": 0.41064453125, | |
| "rewards/rejected": -1.490625023841858, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8226495726495726, | |
| "grad_norm": 12.325756153692222, | |
| "learning_rate": 5.703200127697302e-08, | |
| "logits/chosen": -1.52734375, | |
| "logits/rejected": -1.5070312023162842, | |
| "logps/chosen": -0.526171863079071, | |
| "logps/rejected": -0.6773437261581421, | |
| "loss": 1.2089, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.052343726158142, | |
| "rewards/margins": 0.3001953065395355, | |
| "rewards/rejected": -1.3546874523162842, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 18.975074009248868, | |
| "learning_rate": 5.063687689868209e-08, | |
| "logits/chosen": -1.536718726158142, | |
| "logits/rejected": -1.498437523841858, | |
| "logps/chosen": -0.5103515386581421, | |
| "logps/rejected": -0.698437511920929, | |
| "loss": 1.188, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.596875011920929, | |
| "rewards/chosen": -1.0207030773162842, | |
| "rewards/margins": 0.3758789002895355, | |
| "rewards/rejected": -1.396875023841858, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.844017094017094, | |
| "grad_norm": 10.381047057509468, | |
| "learning_rate": 4.4588854753680106e-08, | |
| "logits/chosen": -1.517187476158142, | |
| "logits/rejected": -1.490625023841858, | |
| "logps/chosen": -0.504101574420929, | |
| "logps/rejected": -0.62890625, | |
| "loss": 1.2251, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.008203148841858, | |
| "rewards/margins": 0.24863281846046448, | |
| "rewards/rejected": -1.2578125, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 10.580168425323338, | |
| "learning_rate": 3.889635341632302e-08, | |
| "logits/chosen": -1.53125, | |
| "logits/rejected": -1.5085937976837158, | |
| "logps/chosen": -0.535937488079071, | |
| "logps/rejected": -0.703125, | |
| "loss": 1.2028, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.598437488079071, | |
| "rewards/chosen": -1.071874976158142, | |
| "rewards/margins": 0.33247071504592896, | |
| "rewards/rejected": -1.40625, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "eval_logits/chosen": -1.384198546409607, | |
| "eval_logits/rejected": -1.3550907373428345, | |
| "eval_logps/chosen": -0.5326991081237793, | |
| "eval_logps/rejected": -0.6943044066429138, | |
| "eval_loss": 1.197023630142212, | |
| "eval_nll_loss": 0.0, | |
| "eval_rewards/accuracies": 0.6236559152603149, | |
| "eval_rewards/chosen": -1.0653982162475586, | |
| "eval_rewards/margins": 0.3231083154678345, | |
| "eval_rewards/rejected": -1.3886088132858276, | |
| "eval_runtime": 26.9387, | |
| "eval_samples_per_second": 72.795, | |
| "eval_steps_per_second": 2.302, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8653846153846154, | |
| "grad_norm": 10.102504447281174, | |
| "learning_rate": 3.356729659201946e-08, | |
| "logits/chosen": -1.5578124523162842, | |
| "logits/rejected": -1.5125000476837158, | |
| "logps/chosen": -0.568359375, | |
| "logps/rejected": -0.699999988079071, | |
| "loss": 1.2354, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.5843750238418579, | |
| "rewards/chosen": -1.13671875, | |
| "rewards/margins": 0.26372069120407104, | |
| "rewards/rejected": -1.399999976158142, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8760683760683761, | |
| "grad_norm": 18.80658014166381, | |
| "learning_rate": 2.8609102087789262e-08, | |
| "logits/chosen": -1.51953125, | |
| "logits/rejected": -1.4835937023162842, | |
| "logps/chosen": -0.532031238079071, | |
| "logps/rejected": -0.6910156011581421, | |
| "loss": 1.2005, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.064062476158142, | |
| "rewards/margins": 0.31684571504592896, | |
| "rewards/rejected": -1.3820312023162842, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8867521367521367, | |
| "grad_norm": 12.388053747492433, | |
| "learning_rate": 2.4028671487010464e-08, | |
| "logits/chosen": -1.509374976158142, | |
| "logits/rejected": -1.489843726158142, | |
| "logps/chosen": -0.525585949420929, | |
| "logps/rejected": -0.6937500238418579, | |
| "loss": 1.19, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.598437488079071, | |
| "rewards/chosen": -1.051171898841858, | |
| "rewards/margins": 0.3369140625, | |
| "rewards/rejected": -1.3875000476837158, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8974358974358975, | |
| "grad_norm": 14.094598355889069, | |
| "learning_rate": 1.9832380542726224e-08, | |
| "logits/chosen": -1.5460937023162842, | |
| "logits/rejected": -1.5265624523162842, | |
| "logps/chosen": -0.535937488079071, | |
| "logps/rejected": -0.703125, | |
| "loss": 1.2081, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.071874976158142, | |
| "rewards/margins": 0.33330076932907104, | |
| "rewards/rejected": -1.40625, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9081196581196581, | |
| "grad_norm": 16.07539311926432, | |
| "learning_rate": 1.6026070302883656e-08, | |
| "logits/chosen": -1.5, | |
| "logits/rejected": -1.463281273841858, | |
| "logps/chosen": -0.521484375, | |
| "logps/rejected": -0.6792968511581421, | |
| "loss": 1.1997, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6156250238418579, | |
| "rewards/chosen": -1.04296875, | |
| "rewards/margins": 0.31560057401657104, | |
| "rewards/rejected": -1.3585937023162842, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9188034188034188, | |
| "grad_norm": 14.515898324276503, | |
| "learning_rate": 1.2615038979858417e-08, | |
| "logits/chosen": -1.540624976158142, | |
| "logits/rejected": -1.5187499523162842, | |
| "logps/chosen": -0.5269531011581421, | |
| "logps/rejected": -0.716796875, | |
| "loss": 1.1778, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": -1.0539062023162842, | |
| "rewards/margins": 0.3780273497104645, | |
| "rewards/rejected": -1.43359375, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9294871794871795, | |
| "grad_norm": 11.346931084998822, | |
| "learning_rate": 9.604034575581698e-09, | |
| "logits/chosen": -1.5398437976837158, | |
| "logits/rejected": -1.5265624523162842, | |
| "logps/chosen": -0.5445312261581421, | |
| "logps/rejected": -0.7339843511581421, | |
| "loss": 1.1779, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -1.0890624523162842, | |
| "rewards/margins": 0.3792968690395355, | |
| "rewards/rejected": -1.4679687023162842, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9401709401709402, | |
| "grad_norm": 9.442038079746466, | |
| "learning_rate": 6.997248272535305e-09, | |
| "logits/chosen": -1.5046875476837158, | |
| "logits/rejected": -1.475000023841858, | |
| "logps/chosen": -0.5355468988418579, | |
| "logps/rejected": -0.689453125, | |
| "loss": 1.1997, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6156250238418579, | |
| "rewards/chosen": -1.0710937976837158, | |
| "rewards/margins": 0.3072265684604645, | |
| "rewards/rejected": -1.37890625, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9508547008547008, | |
| "grad_norm": 15.251874082935148, | |
| "learning_rate": 4.7983085998146596e-09, | |
| "logits/chosen": -1.5398437976837158, | |
| "logits/rejected": -1.5234375, | |
| "logps/chosen": -0.5464843511581421, | |
| "logps/rejected": -0.7046874761581421, | |
| "loss": 1.2014, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0929687023162842, | |
| "rewards/margins": 0.3182617127895355, | |
| "rewards/rejected": -1.4093749523162842, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 16.8382443963138, | |
| "learning_rate": 3.0102763823796506e-09, | |
| "logits/chosen": -1.5203125476837158, | |
| "logits/rejected": -1.4812500476837158, | |
| "logps/chosen": -0.5503906011581421, | |
| "logps/rejected": -0.7093750238418579, | |
| "loss": 1.2004, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.620312511920929, | |
| "rewards/chosen": -1.1007812023162842, | |
| "rewards/margins": 0.3185058534145355, | |
| "rewards/rejected": -1.4187500476837158, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9722222222222222, | |
| "grad_norm": 17.584637370123247, | |
| "learning_rate": 1.6356404805242696e-09, | |
| "logits/chosen": -1.5546875, | |
| "logits/rejected": -1.51953125, | |
| "logps/chosen": -0.5498046875, | |
| "logps/rejected": -0.6880859136581421, | |
| "loss": 1.2288, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.589062511920929, | |
| "rewards/chosen": -1.099609375, | |
| "rewards/margins": 0.27663880586624146, | |
| "rewards/rejected": -1.3761718273162842, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9829059829059829, | |
| "grad_norm": 14.609935824322315, | |
| "learning_rate": 6.763143254950976e-10, | |
| "logits/chosen": -1.506250023841858, | |
| "logits/rejected": -1.4929687976837158, | |
| "logps/chosen": -0.5513671636581421, | |
| "logps/rejected": -0.719921886920929, | |
| "loss": 1.2021, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1027343273162842, | |
| "rewards/margins": 0.33662110567092896, | |
| "rewards/rejected": -1.439843773841858, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9935897435897436, | |
| "grad_norm": 14.658436552359875, | |
| "learning_rate": 1.3363325608131136e-10, | |
| "logits/chosen": -1.5382812023162842, | |
| "logits/rejected": -1.5085937976837158, | |
| "logps/chosen": -0.5210937261581421, | |
| "logps/rejected": -0.701953113079071, | |
| "loss": 1.1787, | |
| "nll_loss": 0.0, | |
| "rewards/accuracies": 0.628125011920929, | |
| "rewards/chosen": -1.0421874523162842, | |
| "rewards/margins": 0.3624023497104645, | |
| "rewards/rejected": -1.403906226158142, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 468, | |
| "total_flos": 0.0, | |
| "train_loss": 1.2459888947315705, | |
| "train_runtime": 3192.5713, | |
| "train_samples_per_second": 18.755, | |
| "train_steps_per_second": 0.147 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 468, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |